Inequalities in the Transition from Good to Not Good Health: Duration Analysis of Spanish Adults

Abstract

Introduction. Healthy time is a valuable good for individuals. Therefore, the longer healthy time lasts, the better. We explore how income and other individual characteristics influence the duration of good health. Our paper constitutes, then, a different way of looking at socioeconomic inequalities in health, using a duration model. Our paper is also related to the literature on health capital (Grossman) as investment in education may also influence the duration of good healh.

Objectives. First, to explore the duration of "good health" and the transition to "poor health" among the Spanish population over 40 years old. Second, to find out the variables that intervene mostly in that duration. Finally, to detect possible differences in the duration of good health by income quintiles and educational level.

Methodology. We have studied the transition from good to poor health using duration models. Our survival time is given by the variable time in good health and the failure variable (event of interest) is defined as the transition from good to poor health.

Given that each individual may fail more than once, we perform two different duration analysis. Only the first failure occurred to each individual (if any) is considered in the first analysis, whereas multiple failures are allowed in the second analysis. The Cox proportional hazard model has been used in both cases. Heterogeneity across individuals and event dependence have been adjusted for a robust estimation of the variance-covariance matrix. Finally, several diagnosis tests are used to test for misspecification.

Data: We use data from the eight waves of the European Community Household Panel (ECHP), which runs from 1994 to 2001. We focus our analysis on the individuals 40 years of age and older whose initial state in the first period is good health, thus keeping 2,129 observations from the full sample.

Results. Regarding the first failure, we find that the richer (upper quintiles), the more educated and the more qualified survive longer in good health than the poorer, the less educated and the less qualified, as it could be expected. On the other hand, suffering a chronic condition (and age in some intervals) have a negative effect on this duration. In the case of multiple failures, the results are similar, although the effects of the explanatory variables are smaller than before.

Conclusions. The variables that intervene mostly in the duration of good health are income, education and having a chronic disease. For example, for the richest quintile the hazard of failure is reduced between 37% and 10% compared to the poorest quintile, depending on the explanatory variables included in the model. Also, as education increases, the hazard ratio of failure goes down. Reducing income inequalities and investing in education should be the priorities to approach the duration of good health in different socioeconomic adult groups.

Keywords: health, duration analysis, transition, income quintile, Spain

JEL Classification: C41, I12